PP2A holoenzymes are heterotrimers composed of one catalytic C subunit (Ca or Cb), one scaffolding A subunit (Aa or Ab), and one regulatory B subunit (B, B', or B''). Compelling evidence that PP2A plays a role in human cancer comes from the finding that the genes encoding the Aa and Ab subunits are mutated in a substantial fraction of human carcinomas, including lung carcinomas. Since lung cancer is mainly due to smoking, which causes mutations, the lung cancer-associated Aa and Ab mutations are likely the result of smoking. Importantly, we demonstrated that all Aa subunits are defective in binding regulatory B or catalytic C subunits or both. The most specific cancer-associated Aa mutations are E64D (lung carcinoma) and E64G (mammary carcinoma). Both mutant proteins are severely defective in binding B'subunits but normal in binding B, B''and C subunits. We hypothesized that B'-containing holoenzymes are tumor suppressors whose formation is prevented by E64D or E64G mutation. To obtain direct evidence for this hypothesis, a knock-in mouse expressing the E64D mutation was generated. In addition, a conditional knock-out mouse, which permits expression of a truncated Aa subunit incapable of binding any B, B', B''and C subunits was engineered. Both mice will be used separately and combined. The specific aims are: 1. To investigate whether homozygous expression of E64D (AaE64D/E64D) leads to increased lung cancer incidence, accelerates lung cancer formation induced by carcinogen, synergizes with a dominant negative mutant of p53, or synergizes with a combination of carcinogen and mutant p53. 2. To investigate whether heterozygous expression of E64D and knockout of the other allele in the lung (AaE64D/-) causes increased lung cancer incidence on its own or in combination with carcinogen and dominant negative p53. 3. To investigate whether knocking out one Aa allele (Aa) or both (Aa-/-) in lung epithelium promotes lung cancer formation. This aim is important because partial or complete elimination of Aa has been observed in various human cancers (breast carcinoma, melanoma, glioma) and may also occur in lung cancer. 4. To prepare mouse embryo fibroblasts expressing E64D in order to investigate how lack of B'-containing holoenzymes affects growth and apoptosis. The effect of B'-holoenzyme loss on the phosphorylation state of key PP2A substrates, including ERK, GSK-3b, BAD, c-Myc, p53 and pRb will be investigated. If E64D expression increases the incidence of lung cancer, we conclude that one or several B'-containing holoenzymes play a role as tumor suppressor in mice and most likely humans. The proposed work establishes a new mouse model for lung carcinogenesis that could be used for drug testing and development. Project Narrative: We hypothesize that a particular form of protein phosphatase 2A (PP2A) is a tumor suppressor. To prove this hypothesis, we propose to generate a mouse model of lung carcinogenesis based on mutations in PP2A that were discovered in human lung cancer. This model could be used for anti-cancer drug testing, as well as drug development facilitated by new insight into the molecular mechanisms that lead to lung cancer.